Accelerometers, tablets and sailing

[Neeves]

I was discussing with a fellow member the problems with measuring speed of yaw at anchor and I thought what I needed was an accelerometer. I Googled 'accelerometer' to discover I already had one. I have an oldish iPad Pro and it has an an inbuilt accelerometer - part of the electronics you pay for, never find and never have a use for.

Paul Rainbow refreshed my memory with this current thread:

Airmar DST810 Triducer with heel, pitch, roll and Bluetooth

to which I posted.

There are a number of apps that use the accelerometer function, SCraMP seems, to me, the best - particularly as it was developed for use in commercial fishing boats. I have looked at 3 Apps in total, and mention the other 2 on my post in Paul's thread. There may be other Apps that are more useful.

I thought to bring it to the wider attention of members, given that Paul thought the Airmar one seemed worth airing. I suspect most people with a tablet don't know (and possibly don't care) that they own an accelerometer but as tablets are recurring thread topic, for navigation, maybe members who have always wanted an accelerometer and did not get one for Xmas would become rapturous

I wonder if members can find the App(s) actually useful to 'us' and why.

Jonathan

 

[AngusMcDoon]

Accelerometers in phones and tablets are used to flip the screen & orientate the camera when you rotate them. If you use that function, you are using them, even if you're not aware that they are there.

 

[Deleted member 36384]

Professor Knox of Knox Anchor fame, developed a strain gauge that recorded in its memory strain over the duration of anchoring. It also had an alarm feature and he developed an algorithm for when the anchor was at risk of dragging. The device can be found by searching google and was once for sale. His algorithm is available on line as well; it may not be his but a formula that was developed previously. I know this is not an accelerometer, but the information from an accelerometer would be used to calculate loads, I assume.

This is a link to a PBO Article that shows Anchorwatch, the device used to measure the forces
https://static1.squarespace.com/sta...02103d72392/1489319853474/JHK+PBO+article.pdf

 

[vas]

had a discussion last year I think on setting up a loadcell and measuring rode forces. Not difficult per se, but cumbersome getting it wired up on deck. However, it may be easier with a longish snubber and fitting the loadcell in a secure waterproof box by the cockpit where one of the snubber mounts are.
Decent lockdown project I recon! anyone?


V.

 

[tjbrace]

Installed SCraMP on my ipad. Love the gyroscope option

 

[ithet]

I have a suspicion that buried deep inside my Raymarine EV-1 sensor core (replacement cost circa £600) is a chip similar to the gyro sensor in my phone (replacement circa £120).

 

[AngusMcDoon]

I have a suspicion that buried deep inside my Raymarine EV-1 sensor core (replacement cost circa £600) is a chip similar to the gyro sensor in my phone (replacement circa £120).

Probably. Cost 72p each if you buy enough, and cheaper still if direct from the manufacturer...

https://uk.farnell.com/nxp/mma8653f...SE-DoublePresence-Jan11-SEMI-ICS-NEWSTRUCTURE

However, paying the computer nerd to write the software to drive it is expensive if you are only selling a low number of products.

 

[vas]

+1000

playing with (well using actually) MPU9250 9DOF for my fin stabilisers s/w, even having drivers and test code is not easy...

 

[Neeves]

had a discussion last year I think on setting up a loadcell and measuring rode forces. Not difficult per se, but cumbersome getting it wired up on deck. However, it may be easier with a longish snubber and fitting the loadcell in a secure waterproof box by the cockpit where one of the snubber mounts are.
Decent lockdown project I recon! anyone?


V.

Try this for size:

Anchor Testing and Rode Loads - Practical Sailor

If its behind a paywall let me know.

You don't want to use a snubber, nothing elastic, simply use dyneema. You need to work all the knots tight and then there is no stretch and it records the tension in the chain rode.

Its not difficult but water is the only problem. I had the load cell at the deck, behind the bow roller. The dyneema was lifted above the actual roller and then attached to the Nast with the load cell 'freely suspended'. I had the display located at the nav table so that I had access to heading, wind direction, wind speed and tension all displayed together. The ideal is to record them automatically all together but that was beyond my skills. The maximum wind speeds, the gusts, were before the maximum tensions. The data I was missing was the speed of the moving yacht (I knew it was moving but could not measure the speed). Scramp now offers that ability but its a devil of a job to set up the measurement system, do so at different scopes and do so with (in my case) increasing wind speeds, from almost calm through to 35 knots.

Jonathan

 

[Neeves]

Installed SCraMP on my ipad. Love the gyroscope option

I have yet to use it in anger. I've tried it in benign conditions, with a glass of wine in hand. I've yet to try it when the wind is singing in the rigging - which is when the extremes of data will occur. But I'd like to collate the data from Scramp with tensions in the rode - and that's a bigger investment in time, see post 9 above.

I conducted my tests in a developing sea breeze which are fairly accurately forecast here in Sydney and develop from near calm, around noon, to as much as 35 knots late in the afternoon. I also chose a location where I had full exposure to wind but wave development was not significant.

I'd like to do it again with Scramp - but need a convergence of conditions (one being it has to be a nice day!)

Jonathan

 

[vas]

Try this for size:

Anchor Testing and Rode Loads - Practical Sailor

If its behind a paywall let me know.

You don't want to use a snubber, nothing elastic, simply use dyneema. You need to work all the knots tight and then there is no stretch and it records the tension in the chain rode.

Its not difficult but water is the only problem. I had the load cell at the deck, behind the bow roller. The dyneema was lifted above the actual roller and then attached to the Nast with the load cell 'freely suspended'. I had the display located at the nav table so that I had access to heading, wind direction, wind speed and tension all displayed together. The ideal is to record them automatically all together but that was beyond my skills. The maximum wind speeds, the gusts, were before the maximum tensions. The data I was missing was the speed of the moving yacht (I knew it was moving but could not measure the speed). Scramp now offers that ability but its a devil of a job to set up the measurement system, do so at different scopes and do so with (in my case) increasing wind speeds, from almost calm through to 35 knots.

Jonathan

thanks J,

I'm slowly working towards integrating all data on NMEA2000 bus and then being able to log subsets over time. Don't hold your breath, but could get something by the summer and try it on my mobo. Not sure about boat speed though, relatively small movements to get anything meaningful from GPS data (considering GPS accuracy as well) unless I'm missing something.

cheers

V.

 

[Neeves]

thanks J,

I'm slowly working towards integrating all data on NMEA2000 bus and then being able to log subsets over time. Don't hold your breath, but could get something by the summer and try it on my mobo. Not sure about boat speed though, relatively small movements to get anything meaningful from GPS data (considering GPS accuracy as well) unless I'm missing something.

cheers

V.

My conclusion was that GPS data for the distances and times needed was useless. This was the motivation for looking at accelerometers and then finding Scramp. I've 'looked' at Scramp and read the background to its development and it might suit. My location for my records were in the shelter of Barrenjoey Headland, on the inshore side. If you have watched the introduction to 'Home and Away' the headland with lighthouse is the introduction to each episode (and most filming is local). I'm anchored about half way along the isthmus over the isthmus from the Tasman Sea (with the swell and surf). Because the headland sticks up a bit the wind comes from both sides as it goes round the obstruction which develops the veering. Veering from wind shear is not uncommon - so it all seemed a legitimate test (or as good as I could get).

The maximum tensions developed were at the end of a veer, hence the focus on determining what was the speed of a veer. Larger tensions are also developed in chop or swell, primarily chop. I tried to minimise the effects of waves, to reduce complications. But maybe with Scramp it would be possible to record seesawing/horsing simultaneously

I wanted the speed as with mass, of our cat, I then can calculate energy. I can measure tension and I can equate our cat's windage to a more commonly owned AWB. I'd then have some crude, but actual, data which would be applicable (with the usual caveats) to predicting equipment qualities - and possibly ending up with a similar answer to the current spread sheets.

The maximum tension I developed was 650kg and the snatch was so violent I terminated the test. A 650kg snatch was frightening but the anchor did not move, I had a stone marking the position of the anchor so could determine if it moved. The anchor has a 'holding capacity' of 3 times that snatch. But anchor drag at much lower tension - which is why I don't believe its simply tension in the rode causing an anchor to drag - there is some other mechanism, but I don't know what it is, I can guess but don't know, and as I don't know I'm not sure how to define nor measure it.

I'm keen for anyone to do some investigation - eventually we might have more answers.

Good luck with your concept - you need lots of patience!

Jonathan

 

[vas]

Ok, so we agree on gps uselessness in this context!
With accelerometers the main issue is sorting out data returned from this tiny chip. I was good at high-school maths and then the little things we did at the uni but not quite the matrix complexity you get from this little 5X5mm chip: linear and polar accelerations on all 3 axis, etc, not fun, not to mention that was 40yrs ago...
Further behaviour will be different on a monohull compared to your cat. I expect your cat with bring higher forces aotbe as less energy will be taken by boat healing.
Which brings us to the second issue that I have to find a suitable local anchorage and persuade the wife that it's OK to anchor in an exposed spot veering left and right

cheers

V.

 

[GHA]

Ok, so we agree on gps uselessness in this context!
With accelerometers the main issue is sorting out data returned from this tiny chip. I was good at high-school maths and then the little things we did at the uni but not quite the matrix complexity you get from this little 5X5mm chip: linear and polar accelerations on all 3 axis, etc, not fun, not to mention that was 40yrs ago...
Further behaviour will be different on a monohull compared to your cat. I expect your cat with bring higher forces aotbe as less energy will be taken by boat healing.
Which brings us to the second issue that I have to find a suitable local anchorage and persuade the wife that it's OK to anchor in an exposed spot veering left and right

cheers

V.

Vas, I still pop in here or a look without logging on once in a while and this one caught my eye
MPU-9250 - way to go!. Openplotter on the Pi i set up for it, it's the recommended gyro for the opensource autopilot. Also has libraries for arduino & esp32. I have one and have had it linked up to openplotter before thought think it might attach/calibrate to the PiPilot app now, though now will have a play using an esp32 writing directly to signalk on a Pi Zero. It has 3 axis accelerometer, 3 axis gyro and 3 axis compass with calibration routines. From an ESP I can write direct to signalk which will then save everything to an influxdb database, for stuff like this a plotting to a database is essential, you'll never get any real use from just looking at the numbers. Easy to set up and use in openplotter.
I've also had a brief play towards rode tension - and old digital kitchen scales looked like far too much fun to throw out so with a HX711 module attached to the load cell it worked straight away, amazingly accurate. For a while I knew the weights of all euro coins to under 1G One day with a few minutes grinding and welding some scrap stainless that will be turned into a rope tension meter, again writing with an ESP either to signalk or direct to influxdb over wifi. Should be interesting though will almost certainly show what we (some of us ) know already about force on the rode - e=1/2mv², try to keep the boat speed down; f=ma, if the boat does move try not to stop it too quickly.

 

[vas]

GHA, glad to see you back!
(or at least lurking and contributing...)
true, my plan was to do all that on my teensies in a similar way.
Just an hour ago managed to get a Raspberry pi 4 (2GB) running Venus OS (Victron) run on the large image with SignalK and node-red. Took me a day to sort out an el-cheapo CAN HAT, but now it works and sees all the NMEA2000 PGNs moving about the bus, brilliant!
Means it should be easy to get the info from any MPU9250 (have one running the stabilisers) and plot the data timestamped with the HX711 load cell.
V.excited on managing the N2K CAN <-> pi link and started studying SignalK now, good thing these lockdowns

will PM you soon my Qs, need to grab the concept a bit more and sort out data and ways to present them.
Task is generally how to show N2K bus data (I select and want) remotely on a browser via the wifi link boat has to the internet.
Then I could even instruct from the browser the N2K bus on the boat to turn on or off the generator, check/log env. conditions, temps, etc...

cheers

V.